Protecting Patients, Preserving Integrity, Advancing Health ......ProtectingPatients,Preserving Integrity,AdvancingHealth: AcceleratingtheImplementationofCOI PoliciesinHumanSubjectsResearch

Protecting Patients, PreservingIntegrity, Advancing Health:Accelerating the Implementation of COIPolicies in Human Subjects Research

A Report of the AAMC-AAU Advisory Committee on FinancialConflicts of Interest in Human Subjects Research

February 2008

Protecting Patients, PreservingIntegrity, Advancing Health:Accelerating the Implementation of COIPolicies in Human Subjects Research

A Report of the AAMC-AAU Advisory Committee on FinancialConflicts of Interest in Human Subjects Research

The AAMC and AAU gratefully acknowledge partial funding for this project fromthe Burroughs Wellcome Fund.

© 2008 All rights reserved.

Table of ContentsExecutive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .v

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

Chapter 1: Policies on Individual Financial Conflicts of Interestin Human Subjects Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5

Chapter 2: Policies on Institutional Financial Conflicts of Interest inHuman Subjects Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13

Chapter 3: Implementation of Conflicts of Interest Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . .22

APPENDIX A:Model Policy on Institutional Conflict of Interest in Human Subjects Research . . . . . . . . . . .36

APPENDIX B:Analyzing Cases Involving Potential Conflicts of Interest in Human Subjects Research:Template and Compendium of Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43

APPENDIX C:Definition of Financial Interests in Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67

APPENDIX D:Institutional Policies and Practices on Consulting: Topics and Question to Consider . . . . . .68

APPENDIX E:Advisory Committee Roster . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70

Protecting Patients, Preserving Integrity, Advancing Health:Accelerating the Implementation of COI policies in Human Subjects Research

v

The vitality and integrity of biomedical research are critical to the health of thepublic and to finding the keys to addressing some of society’s most compellingand difficult challenges. In the United States, universities and medical schools, thedominant source of this research, are now more than ever key components of thesocial, economic and scientific forces that empower our nation in a globalizedeconomy. The academic research community is increasingly aware of pressurescreated by these changed societal expectations, particularly those associated withits relationships with industry. A principled partnership between the academiccommunity and industry is essential if we are to realize the promise of biomedicalresearch, but such collaboration can also create serious conflicts of interest. Thesepressures compel academic institutions to reaffirm their highest values of protectingthe integrity of their research, the well being of the human subjects who participatein it, and the trust of the public.

The federal government has regulated financial conflicts of interest in federallysponsored research since 1995 while giving considerable deference to universitypolicies. Recognizing that this deference rests critically on the trustworthinessand accountability of academic institutions, the Association of American MedicalColleges (AAMC) and the Association of American Universities (AAU) issuedstrong recommendations in 2001 and 2002 addressing individual and institutionalfinancial conflicts of interest in research. In September 2005, in the wake of conflictof interest scandals that rocked the NIH, the Associations organized a workshopof senior university and medical center officials to address the question ofwhether the academic biomedical research community might benefit from moreprecise guidance on how to handle cases where conflicts arise among personaland institutional priorities and values.

Based on the advice received at that workshop, the Associations in 2006 charged anewly formed Advisory Committee with reviewing, and amplifying as appropriate,the Associations’ previously-issued guidance on conflicts of interest in humansubjects research. The Committee was asked to define the key issues confrontingthe biomedical research community, to develop a more nuanced stratification ofthe risks and benefits, to provide detailed guidance about the analysis andmanagement of conflicts of interest, and to provide practical tools to ensure thatacademic research continues to be conducted in conformity with the institutions’core values.

The Advisory Committee, comprised of senior officials at major research universitiesand medical schools, met three times between January 2007 and January 2008;most of its work was accomplished electronically through three working groups.The Committee’s Report addresses the predominant areas of concern in theidentification, analysis, and management of conflicts of interest in human subjectsresearch and presents an array of policy recommendations, as well as educationalguidance and management tools, to assist institutions in dealing more effectivelywith the challenges raised by conflicts of interest. The three chapters of the reportfocus, respectively, on policies on individual conflicts of interest, policies oninstitutional conflicts of interest, and implementation of conflicts managementprograms.

Executive Summary

While these recommendations are refinements of the basic principles expressedin the 2001 and 2002 AAMC and AAU guidance, the Committee strongly advocatesthe adoption of more consistent policies and practices across academic institutions.The Committee also asserts that time is of the essence with respect to fullyimplementing comprehensive conflicts of interest programs in human subjectsresearch.

Chapter 1: Policies on Individual Financial Conflicts of Interest in Human SubjectsResearch

The first chapter, Policies on Individual Financial Conflicts of Interest in HumanSubjects Research, recommends refinements of the existing guidance from theAssociations in three key areas: the definition of covered individuals, the meaningof “compelling circumstances,” and the expansion of reporting and disclosure ofconflicts of interest. Additional recommendations grew out of the Committee’sconsideration of variability of practices in the academic community and concernabout inconsistencies in policies across institutions.

A. Covered Individuals

NIH’s recent Targeted Site Review of 18 grantee institutions revealed that some ofthe institutions’ interpretations of those covered by PHS regulations wereinconsistent with and were narrower than NIH’s interpretation. Accordingly, theAdvisory Committee recommends a definition of “covered individual” that isconsistent with that of NIH.

Recommendation: Based on NIH’s determination that role rather than title shouldcontrol, institutions should adopt in their own conflicts of interest policies the regulation’sbroader interpretation of covered individuals. That is, a covered individual for purposesof an institution’s policy on individual conflicts of interest should be the following: theprincipal investigator and any other person who shares responsibility for the design,conduct, or reporting of funded research, and the spouse and dependent children ofthe investigator and any other person who shares such responsibility.

B. Compelling Circumstances

The 2001 AAMC Report recommended that institutions adopt a rebuttablepresumption against participation in human subjects research by a conflictedinvestigator that could be overcome only by a showing of “compelling circumstances.”Experience with implementing the 2001 recommendation demonstrated a needfor additional explanation of what “compelling circumstances” might mean.

Recommendation: (a) Early-stage research: In addition to the examples ofcompelling circumstances provided in the 2001 AAMC Report, experimentation tofurther develop an early stage discovery may similarly require the insights, knowledge,perseverance, laboratory resources, or special patient populations of the discoverer.The best interests of patients who could benefit from the discovery may justify furtherinvolvement of the discoverer as an investigator. If such circumstances are deemedcompelling by the applicable conflicts of interest committee, the analysis should definethe stages of the research and the specific activities for which there are compelling


vi


vii

reasons for the conflicted discoverer/investigator’s involvement, and an approvedmanagement plan should be structured to restrict the investigator’s roles to thosestages and activities. The management plan should include a clear discussion of thetime line proposed for elimination of the conflicted investigator from researchparticipation and the strategy to restrict the time of involvement of a conflictedinvestigator to a minimum.

Approval and management of the conflict may differ between experiments designedto promote further development on the one hand and those designed to validateclaims linked to the discovery on the other. Approval of such research when humansubjects are involved should require particularly stringent analysis of the degree ofrisk to subjects and of the effectiveness of particular provisions of the conflictmanagement plan to protect subjects and prevent the introduction of bias of theconflicted investigator.

(b) Low risk research: In considering the degree of risk to human subjects, as calledfor in the 2001 Report, an institution’s conflicts of interest committee may encounterhuman subjects studies in which careful assessment finds risk to human subjects tobe sufficiently low that the disposition of any associated conflict by the committeemay be similar or identical to the disposition that would be made by that institutionin non-human subjects research.

C. Reporting of Potential Conflicts of Interest

The Advisory Committee determined that reporting of financial interests byresearchers to their institutions should be clarified and expanded to reduce thepossibility of inadvertent failure to comply with reporting requirements and tobetter assure consistent standards of “relatedness” of reported financial intereststo particular research projects.

Recommendation: The requirement for reporting a covered individual’s outsidefinancial interests that are directly or indirectly related to professional responsibilitiesto the institution should be extended to eliminate any de minimis threshold.

However, the PHS de minimis thresholds may continue to be used to define significantfinancial interest for the purpose of applying the rebuttable presumption againstparticipation by a conflicted investigator in human subjects research.

An institution may wish to consider exempting certain clearly defined types ofconsulting and fees from its definition of reportable financial interests, e.g., fees forserving on grant review committees (study sections), and fees given as honoraria byanother academic institution for an academic activity, such as a seminar or grandrounds presentation.

Recommendation: Covered individuals performing human subjects research shouldbe required to report all of their outside financial interests directly or indirectly relatedto their professional responsibilities to the institution, including their dollar amount,whether or not the individual believes these financial interests might reasonablyappear to be affected by the individual’s current or anticipated human subjectsresearch.

The policy established by the institution should indicate to the reporting individualshow and to what extent their financial information will be handled and shared bythe institution.

D. Pre-Clinical Research

The Advisory Committee believed that certain pre-clinical research may warrantspecial attention where there is a reasonable anticipation of follow-on humansubjects research in the immediate future.

Recommendation:With respect to pre-clinical research, institutions should considerrequiring covered individuals to indicate if their current non-human subjectsresearch that is linked to any of their reportable financial interests is reasonablyanticipated (1) to be a component of an IND submission or (2) to progress to researchinvolving human subjects within the coming 12 months. In such circumstances, theinstitution’s conflicts of interest committee should have the authority to decidewhether any of the policy stipulations that apply to human subjects research shouldapply to this “pre-clinical” stage of the individual’s research.

E. Disclosure of Potential Conflicts of Interest

The Committee emphasized the importance of broadening the scope of disclosureof managed conflicts of interest both inside and outside the institution.

Recommendation: Disclosure should be extended both in scope and in audience, toassure full awareness of potential conflicts and institutional efforts to address them.Specifically, an institution’s policy should require, with respect to any human subjectsresearch project, disclosure of the existence of all financial interests of a coveredindividual that are related to that human subjects research project as follows: to stateand federal officials, as required by statute or regulation; to research funders orsponsors; to all of the researchers, students, and trainees at the institution workingwith the covered individual on the research project in question; to the editors of anypublication to which a covered individual submits a manuscript concerning theresearch; in any substantive public communication of the research results, whetheroral or written; and to the human subjects of the research project, as specified below.Substantive public communication of the research results includes presentations to orinterchanges with the media, and applies whether the audience is lay persons orother professionals.


viii


ix

Institutional policies should provide that the disclosures referenced above shouldgenerally be sufficiently specific to indicate whether the financial interest is anarrangement including but not limited to (i) consulting or other fees, (ii) royalties,(iii) stock, equity or stock options, (iv) an institutionally-defined inventor’s share,(v) a board or other position with advisory or fiduciary duties, (vi) another type ofarrangement, which would be indicated, or a combination of these. Generally, suchdisclosures should also indicate that the conflict has been reported to and is beingmanaged by the institution.

With respect to disclosure to human subjects in research consent forms, and as isindicated in the 2001 AAMC Report, “the precise wording of disclosure in the consentform should be determined by the IRB, but should include an explanation of the factthat the financial interest in question has been reviewed by the COI committee,approved subject to committee oversight, and determined by both the committee andthe IRB not to pose any additional significant risk to the welfare of research subjectsor to the integrity of the research.”

For the purpose of this Recommendation, the document or statement by whichdisclosure is made need not itself contain all the information referenced above, oncondition that the disclosure includes a clear reference to the presence of the conflictinginterest, an indication that additional information is available regarding the detailsof the conflicting interest and how it is being managed, and how that informationcan be readily obtained by those to whom the disclosure is made.

F. IRB Responsibilities Relating to Conflicts of Interest

The Advisory Committee gave special attention to potential conflicts of interestof IRB members.

Recommendation: Institutions should have clear policies, compliant with applicablefederal regulations that address the reporting and management of conflicts of interestof IRB members. The provisions should require reporting of all financial interests(no de minimis threshold) by IRB members, in the same manner as is required forinvestigators, upon their initial appointment to the IRB, with updating annuallyand more often when circumstances change. The provisions should specify how theIRB Chair and/or the Administrator of the IRB will identify and evaluate potentialconflicts of interest of IRB members and make clear that any conflicted IRB membermust be recused from any deliberations relating to studies with which that IRBmember has a potential conflict of interest.

The remaining recommendations in this Chapter, though not specifically focusedon provisions of policies on conflicts of interest in human subjects research,address three issues that the Committee believed merited special attention: inter-institutional variations in policies and practices regarding conflicts of interest,conflicts of interest in clinical practice, and the responsibility of national associa-tions in educating the public on the importance of academic-industry relation-ships and the manner in which conflicts of interest are addressed by academicinstitutions.

G. Institutional Variations in Conflicts of Interest Policies and Practices

Recommendation:While there are clear advantages to having somewhat uniformguidelines and policies for conflicts of interest in clinical research, universities andacademic medical centers vary in many ways, including in their institutional cultures,traditions, missions and objectives, as well as in the populations they serve. TheAdvisory Committee thus endorses the statement on institutional variations in the2001 AAMC Report, but strongly advises against variations that lead to standardsless rigorous than those set forth in this Report.

H. Conflicts of Interest in Clinical Practice

Recommendation: Institutions should adopt policies and establish standards thatminimize bias in the practice of medicine due to real or perceived conflicts of interestof their medical faculty.

I. Educating the Public, the Government, and the Media Regarding Avoidance andManagement of Conflicts of Interest While Supporting Innovation in ClinicalResearch

Recommendation: The Advisory Committee recognizes the benefit to the public ingeneral, and to the patient populations of our academic medical centers in particular,of the translation of research into new therapies, devices, and disease-preventingstrategies. The Committee also recognizes the substantial public benefits that wouldaccrue from improving understanding of the ways in which research is funded, andof the extensive management procedures that are focused on potential conflicts ofinterest in human subjects research in order to avoid or limit any potential adverseeffects from the conflicts.

Therefore, the Committee strongly urges the AAMC, the AAU, and other professionalorganizations to take responsibility for developing and implementing strategies toeducate the public, the media, and the government on the value of academic-industryrelationships, and on how academic medical centers apply their conflicts of interestpolicies to protect the safety of human research subjects and the integrity of clinicalresearch. Such educational strategies need to be implemented on a continuing basisover the long term and should not be confined to episodic responses to instances ofpublic concern.


x


xi

Chapter 2: Policies on Institutional Financial Conflicts of Interest in HumanSubjects Research

The second chapter, Policies on Institutional Financial Conflicts of Interest inHuman Subjects Research, supplements the Associations’ previously-issuedguidance. Institutional conflicts of interest are a source of growing concern as thenumber and complexity of the roles played by institutions of higher educationcontinue to increase, and their relationships with industry continue to expand. Atthe same time, and notwithstanding the earlier guidance documents, the developmentand implementation of comprehensive institutional conflicts of interest policiescontinues to challenge the academic community. Based on this experience, theCommittee addressed many of the especially difficult institutional conflicts ofinterest issues, and its Report offers a template for a policy on institutionalconflicts of interest in human subjects research (Appendix A). The template isnot offered as the model policy but rather as an example of how an institutionmight choose to address the topics identified in this chapter. The Committee’srecommendations follow.

A. Development and Adoption of Policies

Although the Advisory Committee was charged with examining institutionalconflicts of interest specifically in the context of human subjects research, theCommittee urged AAU and AAMC member institutions to commit themselves todevelop and implement comprehensive institutional conflicts of interest policiesthat govern all operational aspects of a university or an academic medical center.Its recommendation on policies addressing institutional conflicts of interest inhuman subjects research is the following.

Recommendation: The Advisory Committee recommends that all AAU institutionsand AAMC schools of medicine and teaching hospitals should:

a. develop an institutional COI policy covering both the financial interestsof the institution and of institutional officials, including deans, departmentchairs and division chiefs, in human subjects research;

b. implement an institutional COI reporting, evaluation, and managementprocess, and create an objective and credible institutional COI reviewprocess involving a standing internal committee or an external reviewentity. Irrespective of their structures, these entities must be empoweredto inform institutional leadership and decision-making;

c. complete policy development and implementation within two years ofissuance of this report.

B. Separation of Administrative Responsibility

The Committee reiterated the key structural element of effective institutionalconflicts of interest programs, management separation, which was initiallyarticulated in the earlier AAMC and AAU guidance.

Recommendation: Research and financial decision-making processes and agentsmust be separated.

C. Rebuttable Presumption

Also carried forward from the earlier AAMC and AAU guidance, this recommen-dation represents the other key principle in responsible institutional conflicts ofinterest programs, that a rebuttable presumption should be established forinstitutional conflicts of interest in human subjects research.

Recommendation: Decisions about whether or not to pursue a particular humansubjects research project in the presence of an institutional conflict of interest shouldbe governed by a “rebuttable presumption” against doing the research at or under theauspices of the conflicted institution.

D. Consistent Implementation

The Advisory Committee believed that it is necessary to the credibility of anyinstitutional conflicts of interest program that the institution and its officials beheld to high standards and that the standards be consistently applied.

Recommendation: Institutions should ensure in policy and practice that institutionalCOIs will be addressed consistently throughout the institution, such that those subjectto institutional financial conflict of interest policies, specifically officials of the institutionand the institutions themselves, are subject to substantive reporting, disclosure, andmanagement of their financial interests to protect the integrity of human subjectsresearch and the subjects who participate in it, as well as institutional values anddecision-making.

E. Institutional COI Committee

This recommendation clarifies that the committee structure that is put into placefor addressing institutional conflicts of interest may be the same as that forindividual conflicts of interest.

Recommendation: Institutions should form a standing institutional COICommittee to review and analyze potential institutional COIs. [Note that institutionsmay choose to use for this purpose the committee established to address individualCOI.] The Committee must be able to analyze when it would be appropriate and inthe public interest to accept and manage a conflict, rather than require that it beeliminated.

All recommendations and considerations pertinent to institutional policydevelopment are incorporated into the template policy. In this manner, theAdvisory Committee hopes to ease the task of developing and implementinginstitutional conflicts of interest policies for those institutions that have not doneso and of refining existing policies for those institutions that have.


xii


xiii

Chapter 3: Implementation of Conflicts of Interest Policies

The third chapter, Implementation of Conflicts of Interest Policies, offers specific,practical advice on putting conflicts of interest policies into practice. The Chapterprovides a detailed discussion of how to analyze conflicts of interest cases as wellas guidance on key elements of conflicts of interest programs. The discussion issupplemented by a short-form template for analysis of conflicts of interest casesand by ten detailed case studies based on real situations at several research intensiveuniversities and medical schools. The cases illustrate the complexity of conflictsof interest and the application of the analysis template to particular fact patterns.The Advisory Committee considers the compilation of cases a work in progressand has asked the AAMC to continue to build upon and update it. TheCommittee commends the use of the cases as teaching tools for probing keyquestions that are implicated when conflicts of interest are present.

The observations and suggestions in this chapter are based on the assumptionthat many inconsistencies and perceived shortcomings across institutions inaddressing research-related conflicts of interest may result from the lack of widelyaccepted templates for doing so. The Advisory Committee believes that adequateeducation of faculty and staff in all aspects of individual and institutional conflictof interest policies and processes is essential to safeguard the integrity and vitalityof the institution and its research and teaching activities, and to assure theprincipled nature of the institution’s relationships with its industry partners. Thefollowing topics are addressed in detail in Chapter 3:

• Analysis of cases involving potential conflicts of interest in clinical research;

• Management of conflicts of interest;

• Monitoring conflicts programs and management of conflicts;

• External professional activities and consulting;

• Education.

The detailed steps presented under these topic headings constitute a model forinstitutions to approach more systematically the analysis and management ofconflicts of interest, and they cumulatively represent a set of recommendationsand standards for responsible and effective conflicts of interest programs. Inaddition, Chapter 3 contains one formal recommendation specifically focusingon education of the research community.

Recommendation: Through its Forum on Conflicts of Interest in Academe, theAAMC should update and supplement the case studies on a continuing basis andmake them available on appropriate websites in order that they may continue to beuseful to the research community as means for teaching about conflicts of interest inhuman subjects research. The AAMC should also make broadly available otherquality teaching tools that have been developed by particular institutions on conflictsof interest.

Conclusion

The Advisory Committee recognized two fundamental problems in the currentstate of affairs in conflicts of interest programs. The first is the lack of consistencyacross academic institutions in the standards for addressing research-relatedconflicts of interest. The other problem involves the tension in the culture oftoday’s academic institutions, where steadily mounting pressure to participate ineconomic development and technology transfer technology may conflict withfundamental academic values. The Advisory Committee believes that its Reportoffers practical ways in which the academic community can address these issuesvigorously, responsibly, and in keeping with their commitment to the integrity ofresearch, the protection of human subjects, and the preservation of public trust.


xiv


1

The vitality and integrity of biomedical research are critical to the health of thepublic and to finding the keys to addressing some of society’s most compellingand difficult challenges. In the United States, universities and medical schools, thedominant source of this research, are now, more than ever, essential parts of thesocial, economic and scientific forces that empower nations in a global world.

Consistent with this new role for academic institutions, the past thirty years havebrought about a major cultural shift in research universities, and especially inmedical schools, with respect to faculty activities. Where once the development ofproducts for the marketplace was discouraged and even looked down upon byresearch faculty, scientists today are encouraged to share their expertise withindustry through consulting, speaking, or other arrangements, to collaboratewith industry in product development, and to form their own companies.

The Bayh-Dole Act of 1980 accelerated this shift by allowing faculty and institutionsto retain title to the intellectual property resulting from their federally supportedresearch and by encouraging them to promote the commercial development oftheir discoveries through technology licensing. There can be no doubt thatBayh-Dole has been an unparalleled success in speeding discoveries from thelaboratory to the marketplace, resulting in great social benefit. Indeed, theDecember 12, 2002 Economist referred to Bayh-Dole as “possibly the mostinspired piece of legislation to be enacted in America over the last half-century.”

However, the benefits of Bayh-Dole and the broader roles of the academiccommunity have come with some potential downsides, and the risks to theintegrity of the research mission of academic institutions and their faculty in thisnew paradigm are decidedly higher. The promises of translational research, thechallenges of technology transfer, and intense expectations at all levels ofgovernment that universities and their academic medical centers function asengines of socio-economic development generate new pressures on institutionsand their faculty members to expand their relationships and deepen theirengagement with industry. These relationships, now encouraged in many forms,may involve financial linkages that are entirely benign but will in other casescarry the potential to create serious conflicts of interest. Moreover, these financialties are occurring in a context of dramatically increased public sensitivity to andconcern with allegations of financial conflicts of interest more broadly in universitybusiness transactions and across diverse sectors of industry.

In recognition of these changing circumstances, both the federal government and theacademic community have been active in defining their respective responsibilitiesto assure the integrity of these commercial relationships. The federal governmenthas regulated financial conflicts of interest in federally sponsored research since1995 while giving considerable deference to university policies. Recognizing thatthis deference rests critically on the trustworthiness and accountability of academicinstitutions, the Association of American Universities (AAU) and the Associationof American Medical Colleges (AAMC) separately and together have beenaddressing conflicts of interest policies and practices for more than two decades.

Introduction

2


In 2001 and 2002, each association issued strong recommendations to its memberinstitutions for developing policies and procedures for the oversight andmanagement of individual and institutional conflicts of interest.1

In response, the academic community has adopted costly programs and proceduresto adapt its policies and practices to these recommendations, as part of a continuingprocess of identifying and acting upon the challenges and opportunities presentedby their relationships with industry. A survey conducted by the AAMC three yearsafter the recommendations were issued confirmed that the academic communityhas taken them seriously, especially those that address individual financial conflictsof interest, and has attempted to bring its institutional policies into conformity.2

With respect to institutional financial conflicts of interest, however, more recentsurvey data indicate that the community continues to struggle with these verycomplex and challenging issues.3

Although these efforts have been substantial and sustained, issues concerningconflicts of interest within the academic research community continue to arise. Arecent report from the NIH of site visits to 18 of its top-ranked awardee institutionsnoted that although the institutions demonstrated a “solid awareness” of theimportance of compliance with federal conflict of interest regulations, some ofthem have not been consistently diligent in timely reporting instances of financialconflicts of interest to the agency as required by the regulations.4 In response tothat report, the U.S. Department of Health and Human Services Office ofInspector General has investigated and sharply criticized the NIH’s performanceof its regulatory oversight responsibilities, and has issued recommendations thatcall for detailed reporting of financial conflicts of interest cases to the agency byawardees. These recommendations underscore yet again how lapses by awardeeinstitutions in fulfilling their commitments in research oversight engenderdistrust and erode confidence in the ability of the community to manage theseissues responsibly on their own.

Integrity of research, the protection of human research subjects, and thepreservation of public trust are paramount values of academic culture. Yet, due totheir deepening engagement with the commercial sector, faculty and institutionsface conflict of interest challenges that were scarcely imaginable even a decade

1 Association of American Medical Colleges, Task Force on Financial Conflicts of Interest in Clinical Research: Protecting Subjects, PreservingTrust, Promoting Progress—Policy and Guidelines for the Oversight of Individual Financial Interests in Human Subjects Research (December 2001).Association of American Universities, Task Force on Research Accountability: Report on Individual and Institutional Financial Conflict of Interest(October 2001). Association of American Medical Colleges, Task Force on Financial Conflicts of Interest in Clinical Research: ProtectingSubjects, Preserving Trust, Promoting Progress II: Principles and Recommendations for Oversight of an Institution’s Financial Interests in HumanSubjects Research (October 2002).

2 Ehringhaus SH and Korn D,USMedical School Policies on Individual Financial Conflicts of Interest. Results of an AAMC Survey (September 2004).

3 Ehringhaus SH,Weissman JS, Sears JL, et al. Responses of Medical Schools to Institutional Conflicts of Interest. JAMA.2008;299(6):665-671.4 National Institutes of Health, Office of Extramural Research: Observations from NIH’s FY 2006 Targeted Site Reviews on Financial Conflict ofInterest, (2/15/2007) http://grants.nih.gov/grants/policy/coi/index.htm.

ago. Accordingly, the associations resolved to examine the extent to which currentconflict of interest policies and management practices in higher education andacademic medicine continue to conform to the institutions’ core values andjustify continued public confidence in the integrity and credibility of academicresearch. The goal of this process is to strengthen and achieve better harmonizationof the policies and practices by which the academic community defines, manages,and oversees potential individual and institutional conflicts of interest in biomedicalresearch.

At least two fundamental problems have been identified in the current state ofaffairs. The first is a lack of consistency across academic institutions in the standardsfor addressing research-related conflicts of interest. While federal regulationsallow considerable institutional flexibility in deference to unique institutionalcultures, such cultural differences should never allow institutions and faculty tostray from the core principles of scientific integrity and protection of humanresearch subjects, or to distort institutional responses to pressures that might biasresearch findings.

The other problem is the mixed message that pervades the culture of today’sacademic institutions. Faculty are acutely aware of the opportunities, incentives,and imperatives presented by the Bayh-Dole legislation as well as steadily mountingpolitical pressure to transfer technology, enhance economic development, andincrease ties between academic medicine and industry. Yet, these forces, while beingacknowledged, must not be allowed to compromise the integrity of their researchand teaching endeavors or encourage the perception that integrity has been com-promised. In accommodating these sometimes competing priorities, institutionalculture must continuously articulate and reinforce the overarching imperative ofprofessional integrity by responsibly monitoring and addressing individual andinstitutional conflicts of interest at all levels of the institution that affect research.

To confront these problems, the AAU and the AAMC convened an AdvisoryCommittee late in 2006 to develop more precise guidance on appropriate responsesto those circumstances in which conflicting personal and institutional prioritiesand values commonly arise. The Committee agreed that review and, as appropriate,clarification and supplementation of the standards articulated in the earlier AAUand AAMC reports would help to strengthen and harmonize institutionalprotections even in the presence of cultural differences. Further, it would enableuniversities and academic medical institutions to weigh better the anticipatedbenefits of specific biomedical research proposals against the risks that specificindividual and institutional financial relationships pose to research participants,scientific integrity, and the institution itself. Finally, such an effort could reassurethe public that the intensifying emphasis on technology transfer and interactionswith industry has not blunted academia’s drive to create and transmit newknowledge based on scientific opportunity and societal need, its commitment toresearch integrity, and its paramount duty to protect human research subjects.

Accordingly, the Advisory Committee has examined the key issues confrontingthe biomedical research community in identifying and managing conflicts ofinterest in research, and, while ascertaining the continued validity of the associations’

3


4


principles and guidance, has recommended refinements and clarifications of theassociations’ previously published statements on these matters.

The report is divided into three chapters and contains a number of appendices.The first chapter, Policies on Individual Financial Conflicts of Interest in HumanSubjects Research, recommends refinements of the existing association guidance inthree key areas: the definition of covered individuals, the meaning of “compellingcircumstances,” and the expansion of reporting and disclosure activities. In addition,it offers an additional recommendation on preclinical research.

The second chapter focuses on Policies on Institutional Financial Conflicts ofInterest in Human Subjects Research and supplements previously issued guidancein this area by AAU (2001) and AAMC (2002), as well as by the Secretary of theDepartment of Health and Human Services (2004). In recognition of the difficultythat many institutions have faced in codifying institutional conflict of intereststandards,5 the report also provides a template for an institutional conflict ofinterest policy (Appendix A).

The third chapter, Implementation of Conflicts of Interest Policies, offers specificguidance on practical aspects of putting conflicts of interest policies into practice.The chapter provides a detailed discussion of how to analyze conflicts of interestcases and provides guidance on the key elements of conflicts of interest programs.This discussion is supplemented by a short-form template for analysis of conflictsof interest cases, and by ten detailed case studies based on real situations at severalresearch intensive universities and medical schools to illustrate the complexity ofconflicts and the application of the analysis template to the fact patterns presented(Appendix B). These case studies will provide an invaluable teaching tool forfaculty, research personnel, students, and other trainees by illustrating concretelythe application of the policy provisions and questions that must be addressed.

The Advisory Committee believes that by developing and adopting nationallyacceptable standards, supplemental guidance, and recommendations for respondingto common conflicts of interest, the academic community should be betterequipped to protect its core principles of research integrity, and to show moreconvincingly that its policies and practices are appropriate to its multi-facetedmissions, and that the steps it takes to protect against conflicts of interest areserious, purposeful, transparent, and effective.

Although this report focuses on those conflicts that arise in the context of humansubjects research conducted primarily within or under the supervision of medicalschools and teaching hospitals or by their personnel, institutions should stronglyconsider making the principles and processes recommended in this reportapplicable to all research. Protection of integrity and public trust are indeed valuesthat underpin all academic research, irrespective of whether the particularchallenges associated with human subjects research are present.

5 Boyd EA, Lipton S, and Bero L. Implementation of Financial Disclosure Policies to Manage Conflicts of Interest. How seven Universities ofCalifornia campuses deal with their research relationships. Health Tracking, March/April 2004.

6 The AAU and AAMC reports are available on the Association websites at aau.edu and aamc.org.

7 National Institutes of Health, Office of Extramural Research: Observations from NIH’s FY 2006 Targeted Site Reviews on Financial Conflict ofInterest, (2/15/2007) http://grants.nih.gov/grants/policy/coi/index.htm.

8 Ibid.

Chapter 1 Policies on Individual Financial Conflicts of Interestin Human Subjects Research

A. Clarification of Individual Financial Conflicts of Interest Policies

The AAMC and the AAU endorse their statements issued in 2001 and 2002regarding policies on individual and institutional financial conflicts of interest.6

Several areas, however, are in need of clarification and refinement. This chapterfocuses on policies on individual financial conflicts of interest (COI).

1. Covered Individuals

Since the adoption of the Public Health Service (PHS) regulations on Objectivityin Research in 1995 (42 CFR Part 50 Subpart F), different practices have developedin the academic community regarding the applicability of the regulations toresearch personnel other than the principal investigator on a PHS-funded project.Those differences became especially apparent during the National Institutes ofHealth (NIH) 2006 Targeted Site Review program, an NIH initiative that focusedspecifically on assessing institutional compliance with the regulations as they pertainto NIH grants. Though the institutions that participated in the reviews werefound to have implemented the regulations “thoughtfully and with diligence,”NIH observed that some institutions define “investigator” too narrowly.7

The regulation defines “investigator” as the principal investigator and any otherperson who is responsible for the design, conduct, or reporting of fundedresearch, and it includes the investigator’s spouse and dependent children (42CFR § 50.603). On the basis of its site review visits NIH reminded institutionsthat the PHS definition of investigator is very broad, and that it is incumbentupon them to determine which individuals are subject to the financial conflict ofinterest regulations. Accordingly, NIH urged institutions to consider the roles,rather than the titles, of those involved in research and the degree of independencewith which those individuals work. “When the definition of investigator is limitedto titles or designations (e.g., to principal investigators, key personnel, faculty)the risk that an unidentified financial conflict of interest may compromise theresearch enterprise increases.”8 NIH has taken the position that the term “coveredindividual” includes the spouse and dependents not only of the principalinvestigator but also of any other person who shares responsibility for the design,conduct, and reporting of research.

5


Recommendation: Based on NIH’s determination that role rather than title shouldcontrol, institutions should adopt in their own conflicts of interest policies the regulation’sbroader interpretation of covered individuals. That is, a covered individual for purposesof an institution’s policy on individual conflicts of interest should be the following:the principal investigator and any other person who shares responsibility for thedesign, conduct, or reporting of funded research, and the spouse and dependentchildren of the investigator and any other person who shares such responsibility.

2. Compelling Circumstances

The 2001 AAMC Report was built on an analytical framework that posited arebuttable presumption against participation by a conflicted investigator inresearch on human subjects. However, that presumption could be overcomeunder certain circumstances so that it would be possible to permit a researcher tocontinue involvement in a research project despite the presence of a clear conflictof interest, where the impacts of the conflict could be adequately managed.

The 2001 AAMC Report provided that:

“In the event of compelling circumstances, an individual holding significantfinancial interests in human subjects research may be permitted to conductthe research. Whether the circumstances are deemed compelling will dependin each case upon the nature of the science, the nature of the interest, howclosely the interest is related to the research, and the degree to which theinterest may be affected by the research.When the financial interest is directlyrelated to the research and may be substantially affected by it, (e.g., an equityinterest in a start-up company that manufactures the investigational product)the risk is greatest and the bar must be high; however, even direct andpotentially lucrative financial interests may be justified in some circumstances.For example, when the individual holding such interests is uniquely qualifiedby virtue of expertise and experience and the research could not otherwise beconducted as safely or effectively without that individual, he or she should bepermitted the opportunity to rebut the presumption against financial interestsby demonstrating these facts to the satisfaction of an institution’s conflict ofinterest (COI) committee. The COI committee might approve the involvementof such an individual in the research, subject to conditions that ensureeffective management of the conflict and credible oversight of the research.”9

Recommendation: (a) Early-stage research: In addition to the examples of compellingcircumstances provided in the 2001 AAMC Report, experimentation to furtherdevelop an early stage discovery may similarly require the insights, knowledge,perseverance, laboratory resources, or special patient populations of the discoverer.The best interests of patients who could benefit from the discovery may justify furtherinvolvement of the discoverer as an investigator. If such circumstances are deemedcompelling by the applicable conflicts of interest committee, the analysis shoulddefine the stages of the research and the specific activities for which there are

9 AAMC Task Force on Financial Conflicts of Interest in Clinical Research: Protecting Subjects, Preserving Trust, Promoting Progress—Policy andGuidelines for the Oversight of Individual Financial Interests in Human Subjects Research, p.7.

6


compelling reasons for the conflicted discoverer/investigator’s involvement, and anapproved management plan should be structured to restrict the investigator’s roles tothose stages and activities. The management plan should include a clear discussionof the time line proposed for elimination of the conflicted investigator from researchparticipation and the strategy to restrict the time of involvement of a conflictedinvestigator to a minimum.

Approval and management of the conflict may differ between experiments designedto promote further development on the one hand and those designed to validateclaims linked to the discovery on the other. Approval of such research when humansubjects are involved should require particularly stringent analysis of the degree ofrisk to subjects and of the effectiveness of particular provisions of the conflictmanagement plan to protect subjects and prevent the introduction of bias of theconflicted investigator.

(b) Low risk research: In considering the degree of risk to human subjects, as calledfor in the 2001 Report, an institution’s conflicts of interest committee may encounterhuman subjects studies in which careful assessment finds risk to human subjects tobe sufficiently low, so that the disposition of any associated conflict by the committeemay be similar or identical to the disposition that would be made by that institutionin non-human subjects research.

Though all of the Cases provided in Appendix B illustrate risk benefit analysisand, to varying degrees, the operation of the compelling circumstances standard,the following represent examples of how the compelling circumstances standardmight be applied under different fact patterns: Case 2, Part B; Case 3; Case 4, PartA; Case 5; and Case 6. Differing levels of risk are illustrated by Case 2, Part A(low); Case 3 (high); and Case 7 (high). How particular management strategiesmight be selected is illustrated by Case 2, Part A; Case 4, Part A; Case 5; Case 7,Part A; Case 9; and Case 10.

3. Reporting of Potential Conflicts of Interest

The AAMC 2001 Report distinguishes between “reporting” and “disclosing” conflictsof interest. “Reporting means the provision of information about significantfinancial interests in human subjects research by a covered individual to responsibleinstitutional officials and to the institutional COI committee, or the transmissionof such information within institutional channels (e.g., from the COI committeeto the IRB).”10 “Disclosure means a release of relevant information about significantfinancial interests in human subjects research to parties outside the institution’sCOI review and management processes (e.g., to research subjects or journaleditors).”11

10 Ibid. p.12.

11 Ibid. p.11.

7


The AAMC 2001 Report advises as follows on the subject of internal reportingwithin the institution:

“Reporting by Covered Individuals. The policy should require covered individualsto report to the institution all significant financial interests that would reasonablyappear to be affected by the individual’s current or anticipated human subjectsresearch. In making such reports, each covered individual should be required todeclare explicitly whether he or she does or does not have such financial interests;the failure to report is unacceptable.

a. Reports should be required at least annually, with prompt updatingwhenever there is an interim, material change in significant financial interests.

b. Some institutions currently require a researcher to indicate on the institu-tional face sheet accompanying the research proposal whether the researcherholds any significant financial interest in the research. All institutions shouldconsider adopting this practice for research involving human subjects.”12

Tracking the 1995 PHS regulation, the AAMC 2001 Report limited reporting bycovered individuals to those financial interests that meet the definition of“significant.” In certain cases, “significant” was tied to a dollar de minimis amountestablished in the PHS regulation of $10,000 (for consulting fees and other kindsof compensation and gifts) and more than $10,000 and 5% ownership interest ina single entity.

Recommendation: The requirement for reporting a covered individual’s outsidefinancial interests that are directly or indirectly related to professional responsibilitiesto the institution should be extended to eliminate any de minimis threshold. (Pleasesee Appendix C for the definition of “financial interests,” as provided in federalregulation but with de minimis amounts deleted.)

However, the PHS de minimis thresholds may continue to be used to define significantfinancial interest for the purpose of applying the rebuttable presumption againstparticipation by a conflicted investigator in human subjects research.

An institution may wish to consider exempting certain clearly defined types ofconsulting and fees from its definition of reportable financial interests, e.g., fees forserving on grant review committees (study sections), and fees given as honoraria byanother academic institution for an academic activity, such as a seminar or grandrounds presentation.

The 2001 AAMC recommendation places a difficult burden on the covered indi-vidual to make the decision about which of his or her outside financial interestswould reasonably appear to be affected by the individual’s current or anticipatedhuman subjects research. This formulation almost certainly creates the potentialfor inconsistencies on the parts of individuals as to which of several interests arereportable. Accordingly, to promote accuracy and consistency, the responsibility

8


12 Ibid. p.17.

for determining which financial interests are relevant to a determination ofconflict of interest should shift to the institution.

Recommendation: Covered individuals performing human subjects research shouldbe required to report all of their outside financial interests directly or indirectly relatedto their professional responsibilities to the institution, including their dollar amount,whether or not the individual believes these financial interests might reasonably appearto be affected by the individual’s current or anticipated human subjects research.

The policy established by the institution should indicate to the reporting individualshow and to what extent their financial information will be handled and shared bythe institution.

This formulation recognizes that individuals may have outside financial intereststhat are unrelated to their professional competence, e.g., that do not relate totheir role as a researcher. It also enables the institution to make accurate andconsistent judgments as to whether a particular financial interest creates a potentialconflict of interest.

Recommendation:With respect to pre-clinical research, institutions should considerrequiring covered individuals to indicate if their current non-human subjectsresearch that is linked to any of their reportable financial interests is reasonablyanticipated (1) to be a component of an IND submission or (2) to progress to researchinvolving human subjects within the coming 12 months. In such circumstances, theinstitution’s conflicts of interest committee should have the authority to decidewhether any of the policy stipulations that apply to human subjects research shouldapply to this “pre-clinical” stage of the individual’s research.

4. Disclosure of Potential Conflicts of Interest

Unlike reporting, which focuses on internal provision of information to the insti-tution about covered individuals’ financial interests, disclosure relates to therelease of relevant information about financial interests in particular human sub-jects research projects to parties outside the institution’s COI review and manage-ment process. The 2001 AAMC Report addresses the issue of disclosure of theexistence of significant financial interests in human subjects research as follows:“The policy should require disclosure of the existence of significant financialinterests in human subjects research as follows: to state and federal officials, asrequired by statute or regulation; to research funders or sponsors; to the editorsof any publication to which a covered individual submits a manuscript concern-ing the research; and in any substantive public communication of the researchresults, whether oral or written.”13

9


13 Ibid. p.18.

It is important to note that while the Advisory Committee recommends thatcovered individuals report to the institution all financial interests related to theirprofessional competence, the Committee recommends that for purpose of disclosurewith respect to any given research project, disclosure is limited to those financialinterests that relate to that research project itself.

Recommendation: Disclosure should be extended both in scope and in audience, toassure full awareness of potential conflicts and institutional efforts to address them.Specifically, an institution’s policy should require that, with respect to any humansubjects research project, disclosure of the existence of all financial interests of a coveredindividual that are related to that human subjects research project as follows: to stateand federal officials, as required by statute or regulation; to research funders or sponsors;to all of the researchers, students, and trainees at the institution working with thecovered individual on the research project in question; to the editors of any publicationto which a covered individual submits a manuscript concerning the research; in anysubstantive public communication of the research results, whether oral or written;and to the human subjects of the research project, as specified below. Substantivepublic communication of the research results includes presentations to or interchangeswith the media, and applies whether the audience is lay persons or other professionals.

Institutional policies should provide that the disclosures referenced above shouldgenerally be sufficiently specific to indicate whether the financial interest is anarrangement including but not limited to (i) consulting or other fees, (ii) royalties,(iii) stock, equity or stock options, (iv) an institutionally-defined inventor’s share, (v)a board or other position with advisory or fiduciary duties, (vi) another type ofarrangement, which would be indicated, or a combination of these. Generally, suchdisclosures should also indicate that the conflict has been reported to and is beingmanaged by the institution.

With respect to disclosure to human subjects in research consent forms, and as isindicated in the 2001 AAMC Report, “the precise wording of disclosure in the consentform should be determined by the IRB, but should include an explanation of the factthat the financial interest in question has been reviewed by the COI committee,approved subject to committee oversight, and determined by both the committee andthe IRB not to pose any additional significant risk to the welfare of research subjectsor to the integrity of the research.”14

For the purpose of this Recommendation, the document or statement by whichdisclosure is made need not itself contain all the information referenced above, oncondition that the disclosure includes a clear reference to the presence of the conflictinginterest, an indication that additional information is available regarding the detailsof the conflicting interest and how it is being managed, and how that informationcan be readily obtained by those to whom the disclosure is made.


10

14 Ibid. p.18.


11

5. IRB Responsibilities Relating to Conflicts of Interest

The following additional provisions should be codified in institutional policy.

Recommendation: Institutions should have clear policies, compliant with applicablefederal regulations that address the reporting and management of conflicts of interestof IRB members. The provisions should require reporting of all financial interests(no de minimis threshold) by IRB members, in the same manner as is required forinvestigators, upon their initial appointment to the IRB, with updating annuallyand more often when circumstances change. The provisions should specify how theIRB Chair and/or the Administrator of the IRB will identify and evaluate potentialconflicts of interest of IRB members and make clear that any conflicted IRB membermust be recused from any deliberations relating to studies with which that IRBmember has a potential conflict of interest.

B. Institutional Variations in Conflicts of Interest Policies and Practices

Under “Core Principles,” the 2001 AAMC Report states: “The Task Force recognizesthat some institutions may determine that additional restrictions are appropriate.Likewise, we do not discourage institutional variations in process or in the allocationof the oversight responsibilities described in this guidance, provided that thereview and management functions that we advocate are performed fully.”15

Recommendation:While there are clear advantages to having somewhat uniformguidelines and policies for conflicts of interest in clinical research, universities andacademic medical centers vary in many ways, including in their institutional cultures,traditions, missions and objectives, as well as in the populations they serve. TheAdvisory Committee thus endorses the statement on institutional variations in the2001 Report, but strongly advises against variations that lead to standards less rigorousthan those set forth in this Report.

C. Conflicts of Interest in Clinical Practice

The Advisory Committee, while respectful of its circumscribed charge withrespect to conflicts of interest in human subjects research, recognizes that manyscientists who engage in human subjects research and have related significantfinancial interests also have active clinical practices in which those financial inter-ests may be problematic and warrant institutional oversight. The Committee alsorecognizes that oversight and management of such conflicting financial interestsof physician faculty in clinical practice settings is warranted.

Recommendation: Institutions should adopt policies and establish standards thatminimize bias in the practice of medicine due to real or perceived conflicts of interestof their medical faculty.

15 Ibid. p.6.

D. Educating the Public, the Government, and the Media Regarding Avoidance andManagement of Conflicts of Interest While Supporting Innovation in ClinicalResearch

Recommendation: The Advisory Committee recognizes the benefit to the public, ingeneral, and to the patient populations of our academic medical centers, in particular,of the translation of research into new therapies, devices, and disease-preventingstrategies. The Committee also recognizes the substantial public benefits that wouldaccrue from improving understanding of the ways in which research is funded, andof the extensive management procedures that are focused on potential conflicts ofinterest in human subjects research in order to avoid or limit any potential adverseeffects from the conflicts.

Therefore, the Committee strongly urges the AAMC, the AAU, and other professionalorganizations to take responsibility for developing and implementing strategies toeducate the public, the media, and the government on the value of academic-industryrelationships, and on how academic medical centers apply their conflicts of interestpolicies to protect the safety of human research subjects and the integrity of clinicalresearch. Such educational strategies need to be implemented on a continuing basisover the long term and should not be confined to episodic responses to instances ofpublic concern.


12


13

Policies on Institutional Financial Conflicts of Interest inHuman Subjects Research

A. Development and Adoption of Policies

Institutional conflicts of interest (Institutional COI), defined as conflicts of interestbased on either the financial interests of the institution itself or of its officials actingin leadership or supervisory positions, are of special concern in the conduct ofhuman subjects research. The perception that research involving human subjects—especially when there is greater than minimal risk for such individuals—might bemotivated even in part by the possibility of personal or institutional financialgain cannot be tolerated. It is one of the core values of academic institutions thatthe protection of research subjects must not be compromised by the existence oreven the reasonable appearance of institutional conflicts of interest.

Many institutions have developed policies and implemented procedures to eliminateor manage institutional conflicts and, at this writing, many others are workinghard to develop and implement such policies. In support of this effort, theAdvisory Committee here provides a reaffirmation, amplification, and refinementof the principles and frameworks already offered in the AAU and AAMC 2001and 2002 guidance documents. Based on the experience of the intervening years,this Chapter discusses some of the especially difficult institutional COI issues thatinstitutions must address, and offers a template for an institutional policy. Thetemplate is not offered as the model policy but rather as an example of how aninstitution might choose to address the topics identified in this chapter (Appendix A).

To the extent institutions have not yet done so, the Advisory Committee stronglyencourages all institutions to develop their own institutional conflicts of interestpolicies consistent with these recommendations. The Advisory Committeebelieves that those institutions that are revising or amplifying their policies mayfind useful the tools offered here. Although the Advisory Committee was chargedwith examining institutional conflicts of interest specifically in the context ofhuman subjects research, the Committee recommends that AAU and AAMCmember institutions should commit themselves to develop and implementcomprehensive institutional conflicts of interest policies that govern all operationalaspects of a university or an academic medical center.

Recommendation: The Advisory Committee recommends that all AAU institutionsand AAMC schools of medicine and teaching hospitals should:

a. develop an institutional COI policy covering both the financial interests of theinstitution and of institutional officials, including deans, department chairs anddivision chiefs, in human subjects research;

b. implement an institutional COI reporting, evaluation, and management process,and create an objective and credible institutional COI review process involving astanding internal committee or an external review entity. Irrespective of theirstructures, these entities must be empowered to inform institutional leadershipand decision-making;

Chapter 2

c. complete policy development and implementation within two years of issuanceof this report.

B. Definition of Institutional Conflict of Interest

There are two basic types of institutional conflicts of interest, one focusing oninstitutional transactions and holdings and the other on activities and holdingsof institutional officials. This duality was well-defined in the 2002 AAMC Report,“Protecting Subjects, Preserving Trust, Promoting Progress II: Principles andRecommendations for Oversight of an Institution’s Financial Interests in HumanSubjects Research”:

An institution may have a conflict of interest in human subjects research wheneverthe financial interests of the institution, or of an institutional official acting withinhis or her authority on behalf of the institution, might affect—or reasonablyappear to affect—institutional processes for the conduct, review, or oversight ofhuman subjects research.16

Understanding the full significance of this two pronged definition is very important.For individuals in high level positions of institutional responsibility, such as adean, department chair, or division chief, a conflict between their personal financialinterests and the institution’s human subjects research is more than an individualconflict; rather, it constitutes an institutional conflict of interest. The potential forinstitutional COI is especially acute in the case of department chair, divisionchiefs, and center and institute directors with financial conflicts of interestbecause they have direct oversight responsibility for individuals doing researchrelated to the conflict.

C. Principles to Guide Management of Institutional COIs

1. Separation of Administrative Responsibility

Recommendation: Research and financial decision-making processes and agentsmust be separated.

If an institution fails to separate its responsibility for the oversight and adminis-tration of human subjects research from its responsibility for the management ofthe institution’s financial interests, the risks of compromising the safety of humansubjects and the integrity of the research performed are significantly heightened.Institutions and institutional officials must segregate individuals who are entrustedwith making decisions about research policy from all decisions, processes, andprojects involving institutional business investments or other financial interests.Most institutions of higher education have established firewalls for these purposes.However, at some high level, the two streams of finance and research oversightinevitably converge. Therefore, systems should be in place, either in the office of


14

16 AAMC Task Force on Financial Conflicts of Interest in Clinical Research: Protecting Subjects,Preserving Trust, Promoting Progress II—Principles and Recommendations for Oversight of anInstitution’s Financial Interests in Human Subjects Research, p.2.


15

the university chancellor or president, or other high ranking university official, toensure that human subjects safety and the integrity of the data generated inhuman subjects research remain the institution’s top priority. Even where theseparation of function is secured by a strong firewall, certain financial relationshipswith commercial research sponsors should be examined closely to avoidinstitutional conflicts of interest.

2. Rebuttable Presumption

Recommendation: Decisions about whether or not to pursue a particular humansubjects research project in the presence of an institutional conflict of interest shouldbe governed by a “rebuttable presumption” against doing the research at or under theauspices of the conflicted institution.

As defined in AAMC’s 2002 Report:

The presumption may be rebutted when the circumstances are compellingand the committee has approved an effective conflict management plan.Whether the [institutional conflicts of interest] committee deems thecircumstances to be compelling should depend in each case upon the natureof the science, the nature of the interest, how closely the interest is related tothe research, the degree of risk that the research poses to human subjects,and the degree to which the interest may be affected by the research. Thecommittee should consider whether the institution is uniquely qualified, byvirtue of its attributes (e.g., special facilities or equipment, unique patientpopulation) and the experience and expertise of its investigators, to conductthe research and safeguard the welfare of the human subjects involved. Evenwhen the institution is deemed uniquely qualified, conflicts associated withsignificant risk to human subjects should be avoided whenever possible and,if permitted, should be managed closely.17

While there may be “compelling” circumstances under which this presumptioncan be overcome, the decision to allow the research to proceed should be madethrough a rigorous, codified and transparent institutional COI evaluation process.

3. Consistent Implementation

The evaluation of whether or not compelling circumstances exist should initiallybe the responsibility of a standing Committee, with ultimate authority for decisionmaking vested in the president/chancellor or governing board, as institutionalgovernance provides. If the institutional COI Committee determines that theproject should be allowed to proceed, it should approve and document the basesand conditions for its decisions, as provided below. If human subjects are involved,the institutional COI Committee report should be provided to the relevant IRB.Under certain circumstances, external IRB review and external monitoring as wellas other tools for managing institutional COI may be considered to ensure thatthe research is adequately monitored. The IRB should require disclosure ofinstitutional COIs in informed consent documents and in other relevant contexts.

17 Ibid. pp. 10-11.

Recommendation: Institutions should assure in policy and practice that institutionalCOIs will be addressed consistently throughout the institution, such that those subjectto institutional financial conflict of interest policies, specifically officials of the institutionand the institutions themselves, are subject to substantive reporting, disclosure, andmanagement of their financial interests to protect the integrity of human subjectsresearch and the subjects who participate in it, as well as institutional values anddecision-making.

D. Key Institutional COI Policy Provisions and Processes

As part of defining its approach to institutional COI, each institution’s policyshould:

1. Determine whether potential conflicts of interest of institutional officials,board members, and officials, including deans, department chairs, divisionchiefs, and center and institute directors, overseeing human subjects researchshould be addressed through the institution’s institutional COI policy or itspolicy on individual conflicts of interest.

2. Determine who the covered institutional officials are. This determinationcould be by title (e.g., Chancellor, President, Dean, Department Chair, Centeror Institute Director, Division Chief, etc.), or by function (i.e., involvement inelements of oversight or decision-making regarding human subjectsresearch), or by both. Institutions should take care to define how inclusivethis designation is to be.

3. Define institution-specific (de minimis or other) reporting standards forinstitutional holdings. In the case of institutional officials, specify which oftheir activities and assets are covered by the institution’s individual conflictsof interest policy, and which fall under the institutional conflicts of interestpolicy.

In addition, institutions should adopt the following processes:

4. All financial interests of the institution and of institutional officials should bereported to a designated institutional office (and updated regularly), andshould be reviewed in accordance with the institutional COI policy.

5. An institutional COI Committee should be established, as is more fullydescribed in Section 7. Note that an institution may choose to use thecommittee that it established to address individual conflicts of interest.

6. The institutional office in charge of technology transfer and licensing shouldreport to the institutional COI Committee (or a designated institutionalofficial) whenever, as a result of a licensing agreement, the institution takes(or intends to take) equity, royalty, or other stake in a sponsor of humansubjects research, and whenever, as part of the agreement, the company ispermitted to participate in human subjects research at the institution.


16


17

7. An institutional reporting link should be established between the institutionalCOI process and the IRB.

8. When the institutional COI Committee determines whether a potentialinstitutional COI should be eliminated or whether it can be managed toallow the research to go forward, such decisions should be documented andcommunicated to affected individuals and officials.

9. Gifts to the University must be differentiated from sponsored researchagreements and grants. A gift should not be burdened with any explicit orimplicit conditions regarding quid pro quo, while a grant may containconditions regarding the sharing of information in a manner consistent withthe university’s mission, research policies, and non-profit status. Both havethe potential to create institutional COI and may need to be considered bythe institutional COI Committee, as provided below.

10. Because gifts and philanthropy can create potential institutional COI,institutions should authorize only designated institutional officials to receivegifts, monetary or otherwise, on behalf of the institution.

11. Institutional COI policies should be publicized on campus and made availableto the public.

E. Identification of Potential Institutional COIs

Given the two pronged definition of institutional COI, two institutional processesare essential to identify potential conflicts. First, systems should be in place toreport to the designated office or official when relationships develop between theinstitution and entities that sponsor research or between entities whose productis being studied in research at the institution. Secondly, systems should be inplace to determine when individuals in a decision-making capacity have financialrelationships that might be perceived to affect their judgment or choices.

One or more of the following circumstances may create potential institutionalCOI in human subjects research. Accordingly, if any one of these exists, theinstitution should conduct a specific, fact-driven inquiry into whether the particularfinancial relationship may affect or reasonably appear to affect human subjectsresearch conducted at or under the auspices of the institution:

1. When the institution is entitled to receive royalties from the sale of theinvestigational product that is the subject of the research;

2. When, through its technology licensing activities or investments related tosuch activities, the institution has obtained an equity interest or an entitlementto equity of any value (including options or warrants) in a non-publicly tradedsponsor of human subjects research at the institution;

3. When, through technology licensing activities or investments related to suchactivities, the institution has obtained an ownership interest or an entitlementto equity (including options or warrants) of greater than $100,000 in value(when valued in reference to current public prices, or, where applicable,using accepted valuation methods), in a publicly-traded sponsor of humansubjects research at the institution; or

4. When, with regard to a specific research project to be conducted at or underthe auspices of the institution, institutional officials with direct responsibilityfor human subjects research hold a significant financial interest in the com-mercial research sponsor or the investigational product. “Significant financialinterest” is defined for this purpose as being consistent with the institution’sindividual conflict of interest policy. In AAMC’s 2002 guidelines for institu-tional COI, the definition includes the following:

a. An equity interest or entitlement to equity (including options or warrants)of any amount in a non-publicly traded company that is i) the sponsor ofhuman subjects research at the institution, or ii) the manufacturer of aproduct to be studied or tested in human subjects research at or underthe auspices of the institution;

b. An equity interest or entitlement to equity (including options or warrants)in excess of the de minimis amount (and not including exceptions forcertain mutual funds), as defined in the AAMC’s 2001 guidelines forindividual financial interests, in a publicly traded sponsor of humansubjects research conducted at or under the auspices of the institution;

c. Consulting fees, advisory board fees, remuneration, honoraria, gifts orother emoluments, or “in kind” compensation from a company that is i)the sponsor of human subjects research at the institution or ii) themanufacturer of a product to be studied or tested in human subjectsresearch at or under the auspices of the institution, that in the aggregateexceed the de minimis amount as defined in the AAMC’s 2001 guidelinesfor individual financial interests, or are expected to exceed that amountin the next 12 months;

d. An appointment to serve, in either a personal or representative capacity,in a fiduciary role for a company that is i) the sponsor of human subjectsresearch at the institution or ii) the manufacturer of a product to bestudied or tested in human subjects research at or under the auspices ofthe institution, whether or not remuneration is received for such service.Typically, the appointment will involve service as an officer, director, orother board member of the company.


18


19

e. An appointment to serve on the scientific advisory board of a commercialsponsor of human subjects research conducted at or under the auspicesof the institution, unless the official has no current significant financialinterest in the sponsor or the investigational product and agrees not tohold such an interest for a period of no less than three years followingcompletion of any related research conducted at or under the auspices ofthe institution.

In addition, there are other relationships that may warrant special scrutiny.Institutions should determine the nature and degree of scrutiny required for anyof these relationships or interests by assessing the potential of institutional COIand weighing the magnitude of any risk to human subjects. Such relationshipsinclude but are not limited to the following:

1. When an investigator, research administrator, or institutional official withresearch oversight authority participates materially in a procurement orpurchasing decision involving major purchases from, or non-routine supplycontracts with, a commercial entity that sponsors human subjects research atthe institution; or

2. When the institution has received substantial gifts (including gifts in kind)from a potential commercial sponsor of human subjects research. Evaluationof the potential sponsor’s gift history might include the following:

a. Whether a gift is of sufficient magnitude that even when held in thegeneral endowment for the benefit of the entire institution, it mightaffect, or reasonably appear to affect, oversight of human subjectsresearch at the institution;

b. Whether a gift is held for the express benefit of the college, school,department, institute or other unit where the human subjects research isto be conducted; or

c. Whether any institutional officer who has the authority, by virtue of hisor her position, to affect or appear to affect the conduct, review oroversight of the proposed human subjects research has been involved inthe solicitation of the gift.

F. Analysis of Potential Institutional COIs

Once a potential institutional COI is identified, a determination must be madewhether or not compelling circumstances justify overriding the rebuttable pre-sumption against conducting the research. The resolution of that question shouldtake into account and document in a report the following:

1. The nature of the science involved;

2. A description of the institutional COI and how closely it is linked to the research;

3. The magnitude of the potential risks to research subjects inherent in the research,and how those risks could be affected as a result

Documents

Protecting Patients, Preserving Integrity, Advancing Health ......ProtectingPatients,Preserving Integrity,AdvancingHealth: AcceleratingtheImplementationofCOI PoliciesinHumanSubjectsResearch